Is managers' generalized trust decreasing? A generational analysis of managers and non-managers using European Social Survey data, 2002–2022

So-called generalized or non-particularized trust, the trust that people have in most other people irrespective of the latter's group affiliations or trustors' prior knowledge of and interpersonal experiences with a trustee (Sztompka, 1999; Uslaner, 2002), is critical for fostering productive social interactions in society (Coleman, 1990; Fukuyama, 1995). However, not just society, organizations also benefit from mutual trust, notably between managers (or principals) and their employees (or agents) (Arrow, 1974; Granovetter, 1985; MacGregor, 1960 [1960]; Kramer, 1999; Tyler, 2003). Managers' trust in other people sustains task delegation and professional specialization while avoiding overly coercive control and expensive and intrusive managerial monitoring and surveillance (McAllister, 1995; Dietz and Den Hartog, 2006). More generally, intra-organizational trust is associated with better organizational performance and employee outcomes (Connell et al., 2003; Covey and Merrill, 2006; Dirks and de Jong, 2022).

Nevertheless, there is widespread concern that generalized trust has been declining. Putnam's book “Bowling alone” famously drew attention to the significant trust decline that has occurred in the United States since the 1960s (Putnam, 2001). Other work has since confirmed this American trend, finding particularly that younger birth cohorts report having significantly less trust in most people than older cohorts do (Clark and Eisenstein, 2013; Schwadel and Stout, 2012; Twenge et al., 2014; Uslaner, 2002). Similar results are reported for many other countries, although among these countries generalized trust may still have risen first and started declining later than in the United States (e.g. Glatz and Eder, 2020; Larsen, 2013). Concerning organizations and their principals, many commentators similarly believe that distrust and excessive managerialism have taken over the workplace (Locke and Spender, 2011; Parker, 2002). Employees are believed to be increasingly stripped of their autonomy (Melman, 2001; Spillane and Joullié, 2022) and subsequently overburdened by compliance and control systems and detached performance metrics (Ajana, 2018; Muller, 2018).

This article brings these two debates on declining trust together. The article's main aim is to present empirical evidence on the possible downward trust trend among managers. In addition, I seek to compare whether and how the generalized trust trend of managers differs from the trend of the general population, i.e. non-managers/employees. If managers' or leaders' generalized trust is steadily declining, this trend should appear as a generational shift (Inglehart, 1971; Mannheim, 1952 [1929]; Ryder, 1965; Tsai and Peng, 2025; Twenge et al., 2010) with younger managerial birth cohorts exhibiting significantly less trust than older cohorts do. As I elaborate below, the idea is that such social change occurs because high-trust older cohorts retire and are succeeded and replaced by low-trust younger cohorts, which is known as generational replacement. The first prediction that I test is that, ceteris paribus, the level of generalized trust of managers declines with their year of birth. Similarly, I compare the trend among managers to the societal trend by assessing how and how much the effect of birth cohort differs between managers and non-managers (employees/the general population). I thereby predict that the trend in generalized trust is more negative among managers than among the general population.

For empirical analysis, I consider the standard indicator of generalized trust, which involves respondents' answer to the question: Generally speaking, would you say that most people can be trusted, or that you can't be too careful in dealing with people (Almond and Verba, 1963; Bauer and Freitag, 2018; Delhey et al., 2011; Rosenberg, 1956; Van Hoorn, 2014). This trust has been used and analyzed by literally thousands of empirical studies since the 1950s/1960s and is widely validated (see Uslaner, 2015 for a review). I define managers as individuals who are responsible for supervising at least one person at work, which highlights the critical role that managers play in shaping employee work experiences and outcomes. Data are large-scale survey data for almost 449,000 individuals for the period 2002–2022. Data are collected in the European Social Survey (ESS), an established database in many social science disciplines.

The empirical results indicate a strong downward generational trend in managers' generalized trust in the sample studied. Moreover, comparing managers and non-managers, it seems that this trend of decreasing trust is twice as strong among the former. This difference, what we may refer to as a managerial generalized trust gap, is the most relevant result because it implies that the downward trend in managers' generalized trust does not merely reflect a broad societal trust decline and that something uniquely affecting managers is happening as well. The results are robust to controlling for a wide range of potential confounders.

Importantly, the results provide an empirical underpinning for the idea that managerialism has gone too far and that there is too much managerial distrust going around in organizations. To wit, absent convincing arguments why younger managerial cohorts are justified in trusting relatively little, the observed managerial generalized trust gap implies that managers' generalized trust is likely suboptimally low ^[1]. At the same time, the downward generational trend has implications for HR management, just as generational shifts in (current and future) organizational members' broader set of values and beliefs have (Ayoobzadeh et al., 2024; Kuron et al., 2015; Ng et al., 2010; Ng and Parry, 2016; Twenge et al., 2010). Following this extant work on the HR implications of generations and generational differences in the workplace, organizations need to prepare themselves for the arrival of a new breed of managers with different beliefs about how much most people can be trusted. A standard practical issue subsequently is how organizations' current HR policies and ways of doing things can be made to fit the generalized trust beliefs of this new generation of managers. However, given the evidence that managers' generalized trust is likely suboptimally low, another salient issue is whether organizations want to appoint and subsequently accommodate low-trusting managers in the first place. Running its natural course, the replacement of higher-trusting managerial cohorts with lower-trusting managerial cohorts inevitably reduces the amount of generalized trust that is going around within an organization. And this then begs the question whether this is a good thing for the organization or not? And, in case of the latter, what mitigating steps can be taken? Given what we know about the positive effects of trust, low generalized trust among managers is particularly undesirable and organizations may thus find it useful to assess (prospective) managers' generalized trust. More generally, employers may implement (additional) trust-building initiatives, such as intensifying onboarding programs and expanding (intergenerational) mentoring relationships between older, established employees and younger newcomers.

This study is situated at the cross-section of different fields. Focusing on managers and trust, the article provides an extension of the literature on generations and generational differences in employees' work-related values (Ayoobzadeh et al., 2024; Kuron et al., 2015; Ng et al., 2010). It further relates strongly to studies of intra-organizational trust and employer-employee relationships and fits the recent plea to bring managers back into management scholarship (Aguinis et al., 2022) and prior work on the prevalence of darker, antisocial or socially aversive mental traits (e.g. psychopathy) among managers (Paleczek et al., 2018; Wille et al., 2013). However, as indicated, the article draws heavily on the large literature on generalized trust that exists outside the HR literature, in, e.g. sociology and political science. A contribution of the article is thus that it brings the concept of generalized trust to the HR literature on trust, which traditionally focuses on so-called particularized trust (Mayer et al., 1995; McAllister, 1995; Vanneste and Gulati, 2022; Vanhala et al., 2011; Yao et al., 2025) ^[2]. In addition to presenting the first comparative analysis of managers' generalized trust, another important empirical contribution of the article is that it brings systematic evidence to the debate about managerialism. As mentioned, an oft-heard idea is that distrust and excessive managerialism have taken over the workplace (Locke and Spender, 2011; Melman, 2001; Parker, 2002; Spillane and Joullié, 2022). However, this debate is not yet infused with systematic and large-scale evidence, let alone with an empirical analysis of trends in managers' generalized trust. Showing that managers' generalized trust is decreasing and that its decline is outpacing the population-wide trust decline, this article helps put this debate on more solid empirical footing.

This article continues as follows. The next section discusses the concept of (generalized) trust and its role in the workplace. The third section elaborates the idea on a downward generational trend in managers' (and non-managers’) generalized trust and formulates hypotheses about these trends. The fourth section presents the empirical approach and data used, while the fifth section presents the empirical results. I end with a discussion and conclusion.

Trust is a multifaceted concept for which many different definitions are available (Bachmann and Zaheer, 2006; Blois, 1999; Chua et al., 2008). Shared among typical definitions is the idea that trust involves a generic willingness to be vulnerable to the actions of others (Kramer, 1999; Mayer et al., 1995; PytlikZillig and Kimbrough, 2016). Within this broad definition, different types of trust can be discerned (Dietz and Den Hartog, 2006; Sztompka, 1999; Uslaner, 2002). An important distinction is between particularized trust and generalized trust (where the latter is also sometimes called impersonal trust) (Nannestad, 2008; Yao et al., 2025). Particularized trust is particular in the sense that it concerns a specific other actor and not people in general. This type of trust – an example is the trust between a manager and a subordinate in their team – is mostly based on prior interpersonal experiences, meaning that it can develop through interactions that take place between the trustor and a known trustee (for example in the workplace). Particularized trust is the type that is most commonly studied in HR research and other organizational literature (Vanneste and Gulati, 2022; Vanhala et al., 2011; Yao et al., 2025). Generalized trust, on the other hand, is less specific and extends beyond personal relationships to people whom one does not know (well). A common definition of generalized trust is as “the belief that most people can be trusted” (Uslaner, 2002, 21). In this article, I consider managers' generalized trust and not their particularized trust such as the trust that they have in their supervisees.

Generalized trust is the essential lubricant that facilitates mutually beneficial exchange between individuals (Arrow, 1974; Fukuyama, 1995; Granovetter, 1985). Generalized trust is thus also essential to organizations as entities in which individuals work together or otherwise interact in ways that are supposed to create value added. Productive or cooperative interactions or exchanges between two actors within an organization, say two team members, involve some degree of vulnerability because the outcome of the exchange for one actor is partly dependent on the behavior of the other actor (Nienaber et al., 2015). Moreover, in many such exchanges, one actor may actually have an incentive to behave opportunistically. If one actor makes the first move, the other actor (i.e. the second mover) often has an opportunity to increase their own payoff by reneging on their promise or not taking the action expected or requested by the first mover. Moreover, recognizing the incentive faced by second-mover actors, first-mover actors may rationally shy away from making the first move altogether. Trust, however, helps overcome this barrier to cooperation and mutually beneficial exchanges (Bijlsma and Koopman, 2003). Generalized trust thereby means that the first mover is not only willing to trust a few specific actors, particularly actors whom they know or have built a relationship with, but many different actors, including actors that they do not actually know and with whom they have no prior interpersonal experiences.

In the workplace, managers' trust in most other people matters in the same way that it matters for social exchanges or interactions in general (Dietz and Den Hartog, 2006). Actors acting as agents on behalf of a manager have room for shirking because monitoring is costly and thus limited and imperfect. Managers, acting as principals, lack the resources and/or expertise to provide (perfect) monitoring of the agents acting on their behalf. As a consequence, these agents – e.g. a subordinate, a fellow manager, or an unknown colleague from another business unit – have leeway in how they do their work, while the manager remains vulnerable to the actions of these actors. Managers thus need to balance the benefits and risks of granting autonomy and delegating tasks to other actors within the organization. However, the more trusting managers are, the more they are willing to forgo extensive and expensive monitoring of actors, which, in turn, sustains task delegation and professional specialization (Meier et al., 2019). Moreover, initial trust by managers may be reciprocated by trustworthy behavior by these trusted actors (Sanders and Schyns, 2006). Distrustful managers on the other hand may overspend on monitoring and control systems and overburden other actors within their organization with compliance paperwork, in addition to failing to reap the full organizational benefits of delegation and specialization.

To be sure, the above discussion is not specific to generalized trust, particularized trust based on trustor's knowledge of or experiences with a specific trustee similarly facilitates productive exchange in the workplace. However, extending beyond specific relationships between a manager and other organizational members, generalized trust has wider relevance for an organization than particularized (manager-supervisee) trust has (Vanhala et al., 2011). Generalized trust entails a broader prosocial disposition that not only concerns team members or other people that one knows personally but also organizational members from other teams, in other functions, in other business units, or even other establishments, at home or abroad. Managers do not always work with or delegate tasks to people whom they know and therefore also need to be willing to trust colleagues whom they have not previously worked with. Any employee who is new to the organization would be one example. The management of remote and distributed workforces (with whom there is often very little face-to-face interaction, if any) is another example. The relevance of managers' generalized trust even extends beyond current employees of an organization to prospective employees. Generalized trust involves being willing to trust a wide group of people, including people outside one's social circle or community (Uslaner, 2002). The higher managers' generalized trust, the more they are willing to trust and hire qualified people who do not belong to their known ingroup (e.g. their extended family), which enables businesses to hire more staff and grow (Akcigit et al., 2021; Ding and Ma, 2024; Fukuyama, 1995; Harrison, 2000; La Porta et al., 1997). Vice versa, low generalized trust is associated with discriminatory tendencies (Harriss, 2003; Van der Linden et al., 2017; Van Hoorn, 2015; Zhang and Wang, 2024), distrusting and excluding qualified people, simply because they do not share the same group relations as one does.

The origins of individuals' generalized trust and differences in generalized trust between individuals can be traced to various factors. A large literature has identified a range of determinants of generalized trust, varying from personal factors such as an individual's genes to broad contextual factors such as countries' formal institutions and (individualistic) culture (see Nannestad, 2008, Schilke et al., 2021; Uslaner, 2002 for reviews). Beyond fixed (biological) determinants of generalized trust, many determinants of generalized trust (within or outside organizations) involve mechanisms and processes by which individuals acquire a certain prosocial trust attitude or willingness to make themselves vulnerable to the behavior of others. Social and institutionalized learning, childhood socialization and parental upbringing are particularly critical. During their life course, individuals gain experience interacting with specific other individuals, learning about trusting and trustworthiness, but are also taught how much they should trust most other people by their family, teachers, peers and community.

Are managers becoming increasingly less trusting, as many commentators have proposed (e.g. Locke and Spender, 2011; Parker, 2002), and, if so, how? Before I present the hypotheses concerning a trend in managers' generalized trust, in this subsection, I first consider the latter question concerning the mechanism underlying systematic change in managers' beliefs such as their trust in most people. Systematic shifts in people's beliefs and values – or social change more generally – have been widely analyzed using the perspective of generations or birth cohorts since at least Mannheim (1952 [1929]) (Alwin and McCammon, 2003; Inglehart, 1971; Ryder, 1965; Tsai and Peng, 2025; see, e.g. Ng et al., 2010; Twenge et al., 2010 for generational studies focusing on work-related mental traits). The logic for taking a generational perspective on the issue of a systematic change in managers' generalized trust comprises four main elements. The first is that birth cohorts share the same socioeconomic and historico-cultural context during their formative years. The second is that this shared or collective context creates shared or collective socialization experiences during childhood, instilling a cohort with a common view of the world that is unique to the cohort and distinguishes the cohort from other cohorts. The third is that these shared pre-adult socialization experiences have a persistent effect, meaning that differences between cohorts are relatively stable over cohorts' life courses. The fourth is so-called generational replacement. Fundamental societal trends occur through the retirement (and death) of older cohorts and the succession and replacement of these cohorts by younger cohorts. Because older and younger cohorts are socialized to have a different view of the world, by way of succession and replacement, the views present in society will change. This change can be systematic: When socialization experiences change successive cohorts' understanding of the world in the same general direction, society as a whole will eventually shift in that direction.

Figure 1 illustrates how demographic succession and cohort replacement may cause a downward trend in the level of generalized trust in a society. Panel A presents the initial situation, three cohorts, with the oldest cohort exhibiting more trust than the middle cohort does, which in turn exhibits more trust than the younger cohort does. The three cohorts together comprise a society where the societal level of generalized trust, shown in the right bar of Panel A, equals the average of the three cohorts ^[3]. Panel B depicts the effect of the oldest cohort dying and being replaced by a newer cohort. The middle and younger cohorts from the initial situation remain. However, because a new cohort with lower generalized trust has replaced the oldest, high-trusting cohort, the aggregate trust level in society goes down. Panel C depicts the effect of further cohort replacement and Panel D the net effect on the level of generalized trust in society, which is sizeable. Overall, systematic differences in the beliefs and values of successive generational cohorts can give rise to fundamental social changes in society.

The main question that this article seeks to answer is whether managers' generalized trust exhibits a downward trend, which is essentially an empirical and not a theoretical question. Following the discussion in the previous subsection, I focus my analysis on generational differences in generalized trust. Generational trends in generalized trust have been widely studied (Clark and Eisenstein, 2013; Glatz and Eder, 2020; Larsen, 2013; Twenge et al., 2014). Moreover, this literature finds robust evidence that in many societies worldwide, generalized trust is steadily declining with each new birth cohort having lower levels of trust in most people than the preceding birth cohort has. Accordingly, we can think of answering the question whether generalized trust is decreasing among managers partly as answering the question whether managers' generalized trust is exhibiting the same downward generational trend observed for the general population. Below, I formulate specific hypotheses on these questions to guide the empirical analysis. However, to be complete, I first take a step back to discuss why generational differences in generalized trust – whether they concern managers or populations as a whole (managers and non-managers) – might occur in the first place.

Following the discussion in the previous section, the overarching driver of generational differences in generalized trust involves differences in the shared socialization experiences that different cohorts have accumulated during their formative years. Theorizing about a society-wide generational trend in generalized trust thus involves considering shared pre-adult socialization experiences that affect cohorts' generalized trust later in life. In general, collective socialization experiences involve being exposed to the social, economic, political, cultural, and technological climates and circumstances of the times during their childhood. The distinction between cohorts is that pre-adult macro-level circumstances have changed. For example, a cohort may have grown up during major events like a (civil) war, a global economic crisis, or political upheaval and other cohorts may not. Similarly, some cohorts have grown up with limited access to electronic media and communication systems whereas others have been fully emerged in digital technologies and the corresponding treatment of public and private information and ways of interacting socially. Still, not all macro-level conditions will end up shaping individuals' generalized trust.

To identify pre-adult macro-level circumstances that likely affect generalized trust, we can take a step back to consider the concepts of trust and generalized trust a bit more. As mentioned, trust involves a trustor's willingness to be vulnerable to the actions of another party, the trustee (Kramer, 1999; Mayer et al., 1995; PytlikZillig and Kimbrough, 2016). This willingness in turn depends on the perceived personal and other-party gains of accepting this vulnerability and an assessment of how risky it would be to trust the other party, but also on trustors' willingness to take risks (Schoorman et al., 2007). Focusing on generalized trust, a further relevant aspect is that the risk assessment and the willingness to take risks concern people in general, not just a limited (in)group of people that the trustor knows and interacts with frequently.

Taking these features of generalized trust, we can consider macro-level developments that likely have affected individuals' pre-adult socialization experiences in ways that have lasting effects on their generalized trust. As mentioned, the idea of a downward generational trend in generalized trust has been much discussed, as well as tested and confirmed empirically (Clark and Eisenstein, 2013; Glatz and Eder, 2020; Larsen, 2013; Putnam, 2001; Schwadel and Stout, 2012; Twenge et al., 2014). Similarly, many studies provide empirical evidence on macro-level determinants of individuals' generalized trust (Nannestad, 2008; Schilke et al., 2021; Uslaner, 2002). What is more, many such macro-level factors can be linked to childhood circumstances that have been changing systematically and would gradually lower trust across cohorts (see, e.g. Freitag and Bühlmann, 2009; Rapp, 2016; Sabatini and Sarracino, 2019; Schmitt-Beck and Wolsing, 2010; Uslaner, 2002). Of course, the number of macro-level developments that could steer pre-adult socialization experiences toward decreasing generalized trust is very large, far too large to analyze in detail in this article. However, let me consider some noticeable macro-level factors and their developments and briefly discuss how they could end up affecting generalized trust, before moving to the conclusion on a generational trust trend among managers.

First, building on Uslaner (2002), rising income inequality may lower consecutive birth cohorts' generalized trust because it divides each specific cohort in different socioeconomic strata with limited intergroup contact, thus limiting the extent to which a strata's trust extends towards other strata in the same cohort (cf. Delhey et al., 2011; Fukuyama, 1995; Van Hoorn, 2014). Rising income inequality may further make individuals more resentful of individuals higher up in the social hierarchy, again limiting the extension of trust toward a wider circle of people and thus lowering generalized trust from one cohort to the next. In addition, and related, inequality may make individuals more prone to zero-sum thinking and less willing to accept that the trusted party may benefit more from the exchange than the trustor themselves do.

Second, building on Rapp (2016), growing (political) polarization may strengthen pre-adults’ ingroup/outgroup thinking as opposed to inclusive thinking. As political identities become more salient and antagonistic, children learn to view political “others” as fundamentally different and potentially threatening, making them unwilling to extend trust toward a wide circle of people that could involve these others. Similarly, polarization frames social relations and interactions as inherently conflictual rather than cooperative, which limits the extent to which people behave prosocially and are willing to engage in an exchange where a potential outgroup member stands to gain (much) from the trustor's trust.

Third, building on Schmitt-Beck and Wolsing (2010), the rise of electronic media and prolonged exposure to these media may nurture perceptions of the world and other people as more dangerous than reality. The reason is that electronic media, particularly visual media, over represent conflict, crime and negative social interactions because of their entertainment value even more than traditional media do (if it bleeds, it leads). Continued pre-adult confrontation with such stories and pictures distorts reality, where most people are in fact not dangerous and can be trusted.

Related, and building on Sabatini and Sarracino (2019), intensifying experience with social media while growing up can also contribute to lower generalized trust later on. Online platforms facilitate constant social comparison through filtered representations of others' lives that foster envious and competitive mindsets rather than prosocial, cooperative ones. Digital environments are further characterized by pervasive bullying and hostility, facilitated by anonymity and the resulting reduced social accountability. At the same time, electronic interactions come at the expense of face-to-face social interaction during critical developmental periods. Hence, overall, younger cohorts will have fewer opportunities to accumulate positive real-world social experiences that build generalized trust.

Finally, building on Freitag and Bühlmann (2009), eroding trust in institutions and politics may instill consecutive birth cohorts with increasingly distrustful demeanors. People in the United States, for example, may have experienced Watergate, with politicians and government showing that it cannot necessarily be relied upon to be impartial and incorruptible. When trusting another party, however, trustors need to know they can rely on formal institutions to help them in case they are cheated. When potential trustors increasingly learn that formal institutions do not necessarily provide such safeguarding, thus increasing the risk of accepting vulnerability to the actions of someone else, the willingness to trust other people will decline.

Of course, to be complete, I should note that there can also be macro-level factors that have evolved in such a way that they would have a positive effect on consecutive cohorts' generalized trust. Indeed, it is impossible for a single article to present an exhaustive theoretical account of all the ways in which socialization experiences can vary and have an impact on the overall generational trend in generalized trust. Taken everything together, however, there is a clear expectation that generalized trust among the general population is, on average, trending downward across cohorts.

Moving on from this conclusion about the general population toward considering this article's main issue of managers' generalized trust, I find no reason to assume that managers are a specific subgroup that is somehow exempted from the downward generational trend in generalized trust established for various populations (Clark and Eisenstein, 2013; Glatz and Eder, 2020; Larsen, 2013; Twenge et al., 2014). Managers have been children just like non-managers have been and have been exposed to the same collective socialization experiences that non-managers have. In fact, the manager/non-manager distinction does not yet exist in individuals' formative years; it is formed much later when some individuals (are selected to) become managers and others are not. Managers and non-managers thus share comparable pre-adult socialization experiences. As a baseline, we can therefore expect that managers' generalized trust also exhibits a downward generational trend. The first hypothesis (H1) correspondingly reads:

Even when managers and non-managers share comparable pre-adult socialization experiences, the debate on the rise of managerialism (Locke and Spender, 2011; Melman, 2001; Parker, 2002; Spillane and Joullié, 2022) gives reason to ask whether there is perhaps something special about managers compared to the general population (i.e. to non-managers/employees). On the one hand, the concern with managers' generalized trust may simply represent the wider concern about declining trust in society (but focused on a specific subgroup). On the other hand, this concern may reflect that there is something special about managers, i.e. about the people who have the same pre-adult socialization experiences that non-managers have but (are selected to) become managers later in life. More generally, a growing number of studies report that certain socially aversive or antisocial mental traits (often labeled dark traits) are more pervasive among managers than among non-managers (e.g. Paleczek et al., 2018; Wille et al., 2013). Generalized trust, in contrast, is inherently prosocial (Uslaner, 2002).

Theoretically, I see two important channels that can sustain a distinct generational trust trend for managers compared to non-managers. The first channel involves the (self-)selection of individuals into a managerial role or not. The second involves managers' and non-managers’ socialization experiences after childhood, particularly in post-secondary education and professional training. Concerning the selection channel, if managers are selected from the general population purely randomly, the individuals that end up being managers will not be systematically different from the individuals that do not end up being managers. In practice, however, managers are selected. Individuals self-select in that some people want to become managers and successfully achieve a managerial position, but not everybody does. Individuals have different life goals and achieving a managerial position often requires certain personal characteristics, including certain mental traits that not everybody possesses. Similarly, organizations select by screening and hiring or appointing some individuals as managers but not others. Managers will thus end up exhibiting a distinct generational trend in generalized trust if these forces of selection have changed in such a way that they correlate with (prospective) managers' level of generalized trust. Among older birth cohorts, high-trust individuals have perhaps been relatively more willing and likely to become managers than low-trust individuals have been, whereas in younger cohorts the reverse holds. Among younger generations, the individuals who are most likely to become managers may also happen to be the individuals who have the lowest level of generalized trust. It is beyond the scope of this study to provide a detailed analysis of how organizations screen and select their managers. However, I find that societal developments provide strong reasons for thinking that selection forces have moved in the direction of more distrust among managers and a stronger downward generational trust trend among managers compared to non-managers. The debate on rising managerialism referenced earlier provides a first pointer for this expectation. This debate reveals a growing concern among a variety of people that, in turn, would be rooted in these people's perceived changes to the workplace. Second, the way work is organized and supervised has been changing fundamentally, specifically the introduction of new technologies and electronic monitoring and control systems, which affects the role and tasks of managers and the traits that they need for their jobs (Höddinghaus and Hertel, 2021; Zuboff, 2019). Compared to the past, where a combination of seniority, loyalty and demonstrated technical skill would be key to becoming a manager, nowadays different qualities are needed, specifically mental traits such as lower trust in people, which fit the prevalence of electronic monitoring and control systems in modern-day workplaces. As the task content of managerial roles has changed, person-job fit has changed, and traits such as low generalized trust that earlier would reduce fit now increase an individual's fit with the role of manager.

Concerning the second channel, manager-specific post-childhood socialization experiences, I expect that business schools and their curricula help create a later-life socialization experience that separates managers from non-managers. The idea is that both current managers and future managers are different from non-managers (i.e. individuals who are not and will never be managers), in part, because the former are more likely to have attended a business school, taken a business course or enrolled in a business- and management-oriented training or tertiary education program. What is more, I find that there have been two relevant systematic shifts in this area, namely in the educational content of business and management programs and in the likelihood that (prospective) managers have followed these programs. Concerning the latter, it seems that managers are increasingly selected because they have had formal education and training in the area of management, with traditional characteristics such as seniority, loyalty and demonstrated technical skill gradually taking a backseat. Concerning the former, there is widespread concern about the theories taught at business schools and how these shape (future) leaders' beliefs about fellow humans. Much of the concern is directed towards the concept of homo economicus, which portrays humans as rational, selfish maximizers, who then require special handling in order to ensure that they are willing to work for the benefit of an organization instead of pursuing their narrow self-interest (Ghoshal and Moran, 1996; Gintis and Khurana, 2008). Two theoretical perspectives, namely agency theory and transaction cost theory, thereby standout for fostering distrust (Ghoshal, 2005): Agency theory suggests that agents cannot be trusted to further a principal's interests unless having formal incentives to do so (Jensen and Meckling, 1976), while transaction cost theory is very much engaged with opportunistic behavior, assuming that this behavior will occur, thus necessitating countermeasures (Williamson, 1975). Over the years, these theories have become part and parcel of business courses and programs, with likely effects on managers' beliefs about people, not least their trust in other people (Hühn, 2014). Different studies indicate that students exposed to economic theories and worldviews become greedier and relatively more self-oriented and less others-oriented and behave less cooperatively and prosocial and more dishonest and antisocial (Etzioni, 2015; Frank et al., 1996; Krishnan, 2008; McCabe and Trevino, 1995; Wang et al., 2011). More generally, there has been concern that the economic theories taught to business students, not least principal-agent theory and its emphasis on incentive alignment and formal contracts as opposed to trust and relational contracts, are becoming self-fulfilling (Ferraro et al., 2005; Pfeffer and Sutton, 2006), promulgating a downward spiral of distrust. Given all this, I expect that managers are indeed distinct from non-managers and experience a generational trust trend that is more negative than the trend among non-managers is. I thus propose the following hypothesis (H2):

I assess the two hypotheses guiding the analysis next.

To analyze whether managers are becoming systematically and fundamentally more distrustful, I take a generational approach. The models that I estimate concern generational cohort or birth cohort as a predictor of managers' (and non-managers’) generalized trust ^[4]. The main challenge when testing for such a generational trend is to separate cohort effects from confounding age and period effects (Hofstede, 2001; Ng and Parry, 2016; Twenge et al., 2010) ^[5]. I follow the cross-classified multilevel approach to such Age-Period-Cohort (APC) analysis proposed by Yang and Land (2006) and also used in Twenge et al.’s (2014) generational analysis of trust among U.S. citizens. This multilevel approach separates variation between individuals at Level 1, the lowest level in the analysis, from variation between periods (years) and variation between age groups at the higher level, Level 2. I further add a standard set of control variables (e.g. individual education and country fixed effects that capture the direct influence of such contextual factors as labor laws and regulations, but also the pervasiveness of violence and crime in society).

Formally, we have the following model for Level 1:

$Giya=β0+β1Ciya+β2Xiya+riy+eiya$ (individual level),

and the following two Level-2 models:

$riy=γ00+u0y$ (year/period group level),

$riy=γ10+u1a$ (age group level).

In these models, G_iya denotes the outcome variable, generalized trust, of individual i in year group y and age group a, C_iya denotes individuals' generational cohort X_iya denotes the control variables such as individuals' educational degree and their years of education, and β₀ is the fixed effects coefficient for the individual-level intercept. r_iy denotes the random intercept for individual i in year group y, while e_iya denotes the residual term for individual i in year group y and age group a. Similarly, γ₀₀ is the fixed effects coefficient for the intercept at the year group level, while γ₁₀ is the fixed effects coefficient for the intercept at the age group level. Finally, u_0y represents the random intercept for year y, capturing the clustering within years, while u_1a represents the random intercept for age group a, capturing the clustering within age groups. Combining the levels, we have a multilevel model that involves intercepts at each level (individual, year and age group), thus capturing all variation in the dependent variable that occurs within year and/or age groups.

The data for this study are cross-country individual-level survey data from the European Social Survey or ESS. The ESS is a biennial survey that has been collecting data on, among others, the attitudes, beliefs and working conditions of individuals living in 32 countries. The first round of the ESS took place in 2002 and the latest available data are for the year 2022, which provides a sufficient time period for estimating cohort effects while taking into account variation between age and year groups. The ESS uses a strict method to ensure that the data collected are comparable across countries and over time.

In eleven rounds (2002–2022), the ESS has collected data for more than 510,000 individuals. However, because of missing data on some of the key variables in the analysis, the main sample that I consider comprises almost 449,000 individuals. This sample covers 32 countries: Albania, Austria, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Estonia, Finland, France, Georgia, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy, Kosovo, Latvia, Lithuania, Luxembourg, Montenegro, Netherlands, North Macedonia, Norway, Poland, Portugal, Romania, Russia and the United Kingdom Most of these countries are located in Europe. Nevertheless, the sample comprises substantial variation on key dimensions of national contexts (Van Hoorn, 2019). The countries in the sample cover six out of the 10 culture clusters identified by House et al. (2004) and seven out of the 12 clusters identified by Hofstede (2001). In addition, the sample comprises countries that are not included in these common culture clusterings (e.g. Ukraine, Lithuania). The sample further comprises well-known transition economies, small open economies and several major economies. Finally, the countries included in the sample represent different varieties of capitalism and welfare regimes as well as chief legal systems in the world (common law, French law, German law, socialist law and Scandinavian law). The ESS is extensively used across social science disciplines and increasingly also in management. The website of the survey, http://www.europeansocialsurvey.org, provides further information.

The dependent variable is a measure of individuals' generalized trust. The ESS includes the standard single-item scale for measuring generalized trust, which has been considered in thousands of studies since the 1950s (Rosenberg, 1956). This item asks respondents whether they belief that most people can be trusted or that one cannot be too careful in dealing with people. The ESS questionnaire item comes with an 11-point Likert-type answer scale that ranges from 0 (You can't be too careful) to 10 (Most people can be trusted). Although this item, or generalized trust more generally, is not much studied in the HR literature, it is widely validated (see Uslaner, 2015 for a survey of the evidence on the indicator's validity and reliability). To be complete, Table A2 in the appendix presents my own evidence on the nomological network and validity of this trust indicator, showing the expected correlations between this indicator and measures of constructs that are logically related to generalized trust such as individuals' perceived fairness of most other people and their trust in specific institutions. Strictly speaking, the measure of generalized trust is an ordinal measure and not a continuous measure. However, I follow the standard approach of analyzing this measure as a continuous measure, which is common for ordinal scales with a large number of points (Norman, 2010; Robitzsch, 2020). Table A1 in the appendix presents details on all the variables in the analysis and their measurement as well as descriptive statistics. Table A6 presents correlations between the main variables in the analysis.

The chief independent variable is individuals' generational cohort. For the main analyses, I operationalize generational cohort as individuals' year of birth. For a robustness check, I also create my own generational cohorts comprising specific birth-year quantiles of the individuals in the sample. The reason for this check is that individuals are not distributed evenly over birth years; there are far fewer individuals in the later and earlier birth years than there are in the middle birth years. Creating my own quantiles, ensures that each cohort comprises almost exactly the same number of individuals. A second important independent variable concerns individuals' managerial status. As indicated, I define managers as individuals who are responsible for supervising at least one person at work, thus paying heed to the idea that managers play a critical role in shaping employee work experiences and outcomes. This definition of managers/non-managers is also used by Van Hoorn (2019) in his generational analysis of managerial values, which uses the same ESS dataset. I use the item that asks respondents whether they are supervising anybody and create a dummy variable that classifies respondents as being a manager when they report yes (1 = manager).

To address biases due to potential confounders, I include several control variables. The key confounders when analyzing a generation trend are individuals' age and the particular period studied (Hofstede, 2001; Ng and Parry, 2016; Twenge et al., 2010). In the analysis, these confounding effects are addressed through the inclusion of period-specific and age-group-specific intercepts (see above). However, I also have baseline control variables that I include in all the models that I estimate. These control variables are variables that other studies identify as predictors of people's generalized trust (see, e.g. Nannestad, 2008, Schilke et al., 2021; Uslaner, 2002) and also commonly included in analyses of cross-country individual-level survey data (e.g. Glatz and Eder, 2020). At the individual level, I control for sex, individuals' educational degree and their years of education. At the contextual level, I include country fixed effects that take into account any biasing effect caused by managers belonging to younger (older) cohorts being more (less) common in some countries in the sample than in others.

For my robustness checks, I consider four sets of additional control variables. The rationale for considering these specific variables mostly comes from my own reasoning and not so much from other studies. In general, the rationale for inclusion is that these additional controls might correlate with individuals' generalized trust and also vary systematically across birth cohorts and with one's managerial status. The first of these sets concerns the type of activities that people do at their work and that managers supervise. The motivation for controlling for the type of supervised activities is that there may be systematic differences in the daily activities performed by non-managers in an organization that may warrant managers to exhibit lower generalized trust. If such systematic differences in the nature of organizational members' daily activities exist, they could, in principle at least, account for any observed generational differences in managers' generalized trust. Practically, I use industry and occupation fixed effects to control for differences in the daily activities performed by non-managers and corresponding differences in the opportunity for shirking, among others. The second set concerns specific individual attitudes and beliefs that might exhibit the same generational trend as the one that I expect managers' generalized trust to have. The motivation for controlling for these variables is the possibility that any observed downward generational trend in (managerial) trust is partly spurious, reflecting a generational trend in one or more other individual traits that in turn correlate with trust attitudes. The specific controls that I add concern religiosity and religious denomination and trust in various institutions, which are often seen as exhibiting a clear generational trend (Shulgin et al., 2019; Twenge et al., 2014). A downside of adding these variables is that they are outcome variables in the same way that trust is an outcome variable, and are thus bad controls (Cinelli et al., 2024). The reason is that the same process that might be causing a generational shift in generalized trust may be causing a generational shift in, say, trust in institutions. And, if so, including these variables may cause an overcontrol bias (Elwert and Winship, 2014). The third set concerns individuals' employment relation (employee, self-employed or working for own family business) and labor market status (paid work, permanently sick or disabled, etc.), which may have a scarring effect on people's generalized trust (Mewes et al., 2021). As for the second set, a downside of adding these controls is that they are partly outcomes of trust so that including them can cause an overcontrol bias. The fourth set concerns individuals' field or subject of study. The motivation for controlling for the field of study is that younger managerial cohorts may have different theoretical knowledge or have been taught different things than older cohorts have, which, in turn, may give rise to systematic trust differences between these two groups (cf. Etzioni, 2015; Frank et al., 1996; Wang et al., 2011). Unfortunately, data on the field of study are only available for Rounds 2–4 of the ESS (2004–2008).

Table 1 presents the results for the baseline test of a generational trend in generalized trust for managers (Model 1) and non-managers/employees (Model 2). There is a strongly negative coefficient for birth year, which confirms both the hypothesized downward generational trend in managers' generalized trust (H1) and the society-wide downward trend reported by other studies (e.g. Glatz and Eder, 2020; Larsen, 2013). Comparing managers and non-managers, the downward generational trend appears to be much stronger among managers than it is among employees, which supports H2. This difference is most relevant because it indicates that the downward trend in trust in most people is partly unique to managers and not merely reflecting a broad trend toward distrust in society at large. Estimates for an interaction term between birth year and managerial status confirm the presence of a district generational trust trend among managers. The coefficient for this interaction term is sizeable and highly statistically significantly negative, providing formal statistical support for H2 (Model 4; see Figure A1 in the appendix for a graphical depiction). Similarly, comparing fit statistics, specifically the -2Loglikelihood, for the model with (Model 4) and without this interaction term (Model 3) shows that adding the interaction term renders a statistically highly significant improvement in model fit.

Reported coefficients for continuous variables are standardized coefficients. Hence, the impact of birth year is that a one standard deviation increase in managers' birth year is associated with 0.047 standard deviations lower trust (−0.047 = −0.023 + −0.024). This effect is smaller than the standardized effect of years of education, which is one of the most robust and strongest predictors of generalized trust. However, it is more than three times larger than the coefficient for sex. The downward generational trend in trust among the general public has been much discussed and is widely seen as problematic since at least Putnam (2001). As is, it seems that any worries about this general trust decline should apply even more to managers, even when superficially a standardized coefficient of about 0.047 may not seem large ^[6].

One possible explanation for the found generational shift is that younger managerial cohorts cannot afford to trust other people too much. In particular, it could be that younger-generation managers are more likely to supervise employee activities and tasks that offer more room for opportunistic behavior and/or in which shirking is more prevalent, which gives them just cause for being less trusting than older-generation managers are. For the first robustness check, I control for people's sector of employment and the specific occupation in which they are working to address any potential effect of such systematic differences in supervisee tasks. The estimated negative coefficient for birth year is a bit smaller than before, both for managers and for non-managers (Model A1 in Table A3). However, the results confirm the baseline results, and I find little support for the idea that systematic differences in the nature of supervised activities can explain the observed decline in generalized trust across managerial cohorts.

For the second robustness check, I assess the extent to which the downward generational trend in (managerial) trust might be reflective of broader generational trends involving other individual attitudes and beliefs. Specifically, I check whether results are robust to controlling for systematic differences in five additional factors. The first two of these concern individuals' religiosity and denomination, which may exhibit a strong trend towards secularization. The other three factors concern individuals' trust in specific institutions, namely the police, the legal system and politicians, which is widely considered to be declining as well. Adding these five controls renders a negative coefficient for birth year but the estimated coefficient is smaller than before, particularly for non-managers (Model A2). For managers, the generational shift is also less strong than before but still strongly and significantly negative (−0.005 + −0.022 = −0.027). Results thus do not support the idea that the observed decline in generalized trust across managerial cohorts is spurious. Meanwhile, concerning effect sizes, it is interesting to note that the coefficient for religiosity, another well-known predictor of trust, is smaller than the coefficients for birth year reported in Table 1.

For the third robustness check, I assess the extent to which systematic differences in the employment status of older and younger managerial cohorts may account for the observed generational differences in managers' generalized trust. Older managers may be less likely to be self-employed and more likely to be working for a family firm, for instance, and these differences may have an impact on the found generational trends. However, results with controls for employment status added (Model A3) are similar to the baseline results.

For the fourth robustness check, I assess whether systematic differences in field of education of older and younger managerial cohorts may account for the observed generational differences in managers' generalized trust. The baseline model controls for differences in educational degree and years of education. However, older and younger managerial cohorts may have different trust attitudes because they are educated differently and have a different understanding of what the supervision of people entails. Model A4 presents the results. Although these results pertain to a relatively small sample that covers only the years 2004, 2006 and 2008, they confirm the baseline results.

For the fifth robustness check, I consider a potential biasing effect caused by cross-country heterogeneity in the effects of age and year on generalized trust. As indicated, assessing a generational trend requires separating cohort effects from confounding age and period/year effects. I have done so by including random intercepts for both year groups and age groups simultaneously (Yang and Land, 2006). A limitation thereby is that the effects of age and year on trust are implicitly seen as the same for all the countries in the sample. To address this limitation, I repeat the baseline analysis but with individuals cross-classified in country-specific year groups and country-specific age groups (Level 2) instead of in year groups and age groups. Results (Model A5) are almost exactly the same as in the baseline analysis. Because model fit improves highly statistically significantly compared to Model 4, it seems that there is indeed significant country heterogeneity in the effects of age and year on managers' and non-managers’ generalized trust.

For the sixth robustness check, I assess whether the baseline results may be driven by individuals with extreme scores on birth year and age. To do so, I limit the sample to comprise individuals scoring between the 1st and 99th percentiles of the age and birth year variables. Excluding the outliers below and above these percentiles renders estimates that are very close to the estimates from the baseline analysis (Models A6 and A7 in Table A4).

For the final robustness check, I assess whether the baseline results are sensitive to different operationalizations of the concept of generation or generational cohort. Some birth years may comprise many more individuals than other birth years. Hence, I also constructed two alternative measures of individuals' generational cohorts based on birth-year ventiles (with cohorts numbered 1–20) and birth-year fortiles (with cohorts numbered 1–40). Results (Models A8-A11 in Table A4) reveal the same distinct generational trends as before, indicating that the baseline results are robust and not sensitive to considering different types of generational indicators.

As a second extension I consider how the trust levels of different managerial cohorts compare to the level of trust of the general population. The baseline analysis indicates that managers are, on average, slightly more trusting than the general population is (Models 3 and 4 in Table 1). This result is in line with prior empirical work (Carmo and Nunes, 2013). At the same time, the stronger downward generational trend among managers compared to non-managers indicates that this difference is disappearing and reversing into a managerial generalized trust gap (see, also, Figure A1).

Table A5 presents the estimates for a model that compares the trust levels of younger and older managerial cohorts with the trust level of the general population. I thereby define the younger managerial cohort as managers born after 1980 (which is the birth year that is one standard deviation above the average birth year in the sample). As expected, results indicate that these younger-generation managers are significantly less trusting than the general population is, whereas older-generation managers (born before 1981) are indeed significantly more trusting (Model A12). This result holds while taking into account the direct effect of people's birth year on their trust level.

Is managers' generalized trust decreasing? To answer this question, I have systematically compared managers' level of generalized trust across different birth years or generational cohorts. The results provide strong evidence that generalized trust is significantly higher among managers belonging to older birth cohorts than among younger cohorts. Results indicate that a similar downward generational trust trend exists among non-managers. However, the downward trend is about two times stronger among managers than among non-managers. This result suggests that younger managerial cohorts will have lower generalized trust than their supervisees have, which is confirmed by further evidence. Results are robust to the inclusion of various additional control variables. Three main findings are thus as follows. First, both managers and non-managers are becoming systematically less trustful of most people. Second, managers' generalized trust is declining even more across cohorts than the generalized trust of the general population is. And, combining the two previous findings, the third main finding is that there is a generalized trust gap with younger managerial cohorts trusting less than their non-managerial counterparts do.

The above findings contribute insights to various fields and literature. Extending prior generational studies (Ayoobzadeh et al., 2024; Kuron et al., 2015; Ng et al., 2010), they show that generational differences extend beyond work-specific beliefs and attitudes to generalized trust as another mental trait with critical workplace implications. More generally, this article of course, helps raise attention for the concept of generalized trust, which is highly relevant to HR (Vanhala et al., 2011), but has traditionally taken a backseat to the concept of particularized trust.

A second important insight concerns the debate about managerialism. Although the idea that distrust and excessive managerialism have taken over the workplace is much discussed (Locke and Spender, 2011; Melman, 2001; Parker, 2002; Spillane and Joullié, 2022), this debate so far has lacked systematic and large-scale evidence. Fitting this debate, the results of the present study suggest that a substantial shift may indeed be taking place in the workplace. New organizational forms and technologies, specifically information and communication tools, may affect the degree to which employees are trusted as opposed to subjected to intensive monitoring and control (Höddinghaus and Hertel, 2021; Zuboff, 2019). Outside the HR and organizational literature, there have been many studies discussing and empirically testing a society-wide decline in generalized trust (Clark and Eisenstein, 2013; Glatz and Eder, 2020; Larsen, 2013; Putnam, 2001; Schwadel and Stout, 2012; Twenge et al., 2014). This article's findings, particularly the strength of the downward generational trend among managers and the resulting generalized trust gap between younger managerial cohorts and their non-managerial counterparts, also adds to this long-standing debate. There is sizeable heterogeneity between groups in society and managers form a group that is particularly noticeable for its strongly negative generational trust gradient.

The article further adds insight to the body of research that considers mental traits that distinguish managers from non-managers and affect individuals' rise through organizational ranks. This literature reports significant differences in mental traits between managers and non-managers, particularly in the areas of narcissism, Machiavellianism and psychopathy, which are socially aversive, dark traits (Paleczek et al., 2018; Wille et al., 2013). The added insight is that managers may also differ from non-managers in that their prosocial attitudes such as generalized trust are declining more rapidly, which is implied by the finding that the downward generational trend in generalized trust is twice as strong among managers compared to managers. More broadly, although the article does not test specific mechanisms or channels, the distinct results for managers vis-à-vis non-managers are consistent with the idea that something may have changed in the way organizations select managers or in the kind of mental traits that are required for managers in modern-day workplaces. That is, the way in which work is organized nowadays may, through different channels, play a role in the unique generational decrease in generalized trust that we observe among managers.

Related to the points discussed in the last two paragraphs, the study's findings also fit the debate on business schools and their curricula. Although the evidence is circumstantial, the idea is that business and management programs may have contributed to managers' distinct generalized trust decline and the managerial generalized trust gap. Business schools are often criticized for their effect on (prospective) managers' beliefs about fellow humans, not least the latter's selfish motivations, which would undermine the former's trust in the latter. At the same time, individuals are increasingly selected for managerial roles because they have had formal education and training in the area of management, thus slowly expanding the impact that business schools' teachings have on younger-generation managers' generalized trust.

To conclude, one final insight concerns the suboptimality of younger-generation managers' generalized trust. The evidence on the generalized trust gap between younger managerial cohorts and their non-managerial counterparts suggests that managers have, in fact, become overly distrustful (which, in turn, would harm organizations and their members). Purely theoretically speaking, younger-generation managers may be justified in trusting significantly less than comparable non-managers do. Indeed, I cannot readily identify the exact level of a manager's generalized trust that is “optimal” for their organization. However, there are no obvious reasons that justify the observed managerial generalized trust gap. In fact, the opposite is much more plausible, meaning that younger managerial cohorts would actually be justified to trust most people at least as much as older cohorts and/or non-managers do. Notably, Mastroianni and Gilbert (2023) find that there is no decline in the degree to which people are kind, honest, nice and good, which in turn implies that there is no evidence of an increase in misbehavior that would justify younger managerial cohorts to trust most people less. What is more, there is growing evidence that people overestimate the degree to which individuals other than themselves cheat and behave dishonestly. Martuza et al. (2024), for example, conducted experiments showing that subjects overestimate other people's dishonesty by some 14% points on average (see, e.g. Alper et al., 2024, Riedl et al., 2025; Teper et al., 2015 for similar evidence). Hence, there is not only evidence indicating that trusting less is not justified but even evidence that trusting more is justified. Interesting, this latter conclusion is itself supported by work that explicitly considers whether people trust too much or too little. Butler et al. (2016), for example, find that trusting too little is far more costly to individuals than trusting too much (see also Lange, 2015). Hence, the conclusion that, despite the absence of direct evidence, the fact that the trust level of younger managerial cohorts is significantly lower than the trust level of comparable non-managers indicates that these managers have suboptimally low levels of trust.

In broad terms, the downward generational trend in generalized has the kind of implications for HR management considered extensively in the literature on generations in the workplace (Campbell et al., 2015; Ng et al., 2010; Ng and Parry, 2016; Twenge et al., 2010). Organizations thus need to prepare themselves for the arrival of low-trusting organizational members (managers and non-managers alike) as well as the retirement of older, high-trusting organizational members. The finding that the generalized trust of younger-generation of managers is likely suboptimally low thereby adds a dimension to the standard issue of how organizations' current HR policies and ways of doing things can be made to fit a new (managerial) generation. Concerning the practical implications of the study's findings, it therefore helps to distinguish between the generic implications of the found generational trust differences in general, concerning all younger-generation organizational members irrespective of whether they are managers or not, and the managerial generalized trust gap in specific.

For thinking about how organizations can prepare themselves for and particularly adapt their HR policies to a new, low-trust generation, it helps to consider again what it means for people to have low generalized trust. Generalized trust involves the belief that most people can be trusted. To have low generalized trust thus means that (1) someone does not trust most people (only a few), or that (2) one has little trust in most people, or (3) both. Both trusting a few people and having low trust in people in general have consequences for organizations. In addition, individuals' generalized trust matters not only when they are full-fledged members of the organization but also earlier, at the recruitment stage.

To start with the latter, an implication of low generalized trust is that younger-generation workers will, on average, be more skeptical towards and less trusting of an organization and the people that it employs, including recruiters and other HR professionals. They may thus invest more in researching a (prospective) employer and checking whether reality matches its public image and the information that they received from their organizational contacts. Similarly, onboarding at the start of the employment relationship may require more focus on showing that managers, senior staff and direct colleagues – though not yet well known to the new hire – can be trusted. In general, onboarding processes need to pay more attention to blurring the boundary between the new hire and an established team quickly, thus avoiding the prolonged existence of an ingroup/outgroup division between the team and the hire and corresponding distrust.

From thereon, the need for additional trust building continues. In addition, as full-fledged members of an organization, younger-generation workers would on average be sensitive to trust issues. Senior management in particular should thus act in a reliable, consistent way, showing their younger-generation supervisees that they can take a chance and make themselves vulnerable to managerial behaviors, actions and decisions. Transparency and competence seem especially valuable here (Klimchak et al., 2020). Interpersonal relationships remain similarly critical at this stage of the employment relationship, as they help ensure that an employee feels part of a wider organizational group and not an outsider who feels the need to be careful when dealing with other organizational members.

Loosely building on the intra-organizational trust-building and trust-repair literature (see, e.g. Kähkönen et al., 2021, Kramer and Lewicki, 2010, Kong et al., 2025; Lewicki and Brinsfield, 2017 for reviews), some examples of specific trust-building initiatives that organizations, including HR departments and professionals, could implement are as follows. The first is to grant employees more autonomy and ownership over their work. This signals trust in one's employees, which could be reciprocated in trust toward a broad group of other organizational members (Sanders and Schyns, 2006). The second is that employees from different generations may be paired, perhaps in mentor-mentee relationships. Such cross-cohort pairing facilitates the breaking down of generational faultlines that otherwise promote ingroup/outgroup thinking and limit mutual trust. More generally, more attention may be paid to group and team building, including with members from other business units or departments that an employee does not regularly interact with. Creating a minimum of prior interpersonal contact may help younger generations reach out to colleagues beyond their immediate circle should their activities take them to other, lesser-known parts of their organization. Recognition and appreciation can be further made more explicit. Consistently acknowledging and rewarding employee contributions to demonstrate that they are valued as individuals helps build the kind of relationships that foster trust. Finally, though perhaps not an initiative strictly, it seems wise for organizations to develop a system that can track employees' generalized trust levels. The resulting data can be used to assess the effectiveness of (intergenerational) trust-building initiatives. However, it may also be worthwhile to assess (prospective) employees' and new hires' generalized trust as part of an individualized onboarding and development program.

Many of the implications discussed above, as well as the recommendations made, are also applicable when focusing on the population subgroup of younger-generation managers. However, the things discussed above gain relevance because managers play such pivotal roles in organizations and because managers' generalized trust levels appear to have reached a suboptimally low level. More than employees, managers shape what is going on in organizations, including how much trust there is. Compared to employees, managers are, for example, more involved in recruitment and selection of new organizational members, which is significantly affected by generalized trust (Akcigit et al., 2021; Zhang and Wang, 2024). Hence, more than the downward trust trend among the general population, the downward managerial generalized trust trend is something that HR policies in organizations would need to reckon with.

Concretely, all the implications and recommendations mentioned above are applicable to managers as well. Younger-generation managers would also benefit if senior leadership acts and behaves in a consistent and reliable way or from intergenerational mentor-mentee relationships. An older-generation manager could, for example, mentor a younger-generation manager, with a lot of emphasis on establishing early connections and interactions with a broad group of people in the organization, breaking down initial barriers to trust. However, because managers and their generalized trust are so important for organizations, efforts in this area deserve to be even more intense than for employees in general. I thus suggest that preparing for the arrival of a new breed of comparatively distrusting managers cannot get around to setting up a system for assessing and monitoring these managers' generalized trust. Lacking trust is perhaps not a dark trait but being overly distrustful of most people is still not a desirable managerial trait. Hence, a concrete recommendation is to pay attention to managerial candidates' level of generalized trust early on already, when recruiting and selecting individuals for leadership positions. Although not without its limits, a formal, non-clinical generalized trust assessment could be part of the selection, akin to the idea of screening for dark traits among managers (Judge and LePine, 2007; Reichin et al., 2019). From thereon, individuals who are selected and appointed can be tracked and their generalized trust nourished further using targeted initiatives and interventions along the lines discussed above (see, also, Kähkönen et al., 2021, Kramer and Lewicki, 2010, Kong et al., 2025; Lewicki and Brinsfield, 2017).

The general robustness of this article's findings notwithstanding, there are also some limitations to the analysis presented here. The main limitation of my analysis is that I have not considered specific factors and/or mechanisms that would explain the main findings. For example, I have not studied specific generational differences in socialization experiences. In addition, and even more relevant to the HR and organizational literature, this study has not taken a close look at the process by which some individuals self-select and/or are selected to become managers and others are not. I have touched upon this issue, but the discussion was necessarily brief. Nevertheless, because employer-driven selection and self-selection cause some individuals in the sample to be managers and others to be non-managers/employees, these selection forces are likely at the root of the distinct generalized trust trends found as well as the resulting managerial generalized trust gap. Still, we do not yet know exactly what sort of selection forces could be underlying the strongly negative generalized trust trend among managers and the resulting managerial generalized trust gap. Is it some form of self-selection or (undeliberate) employer-driven selection or a combination of the two? Considering the perspective of individuals self-selecting into (non-)managerial roles, it might be that becoming or being a manager is not as attractive as it once was, which has resulted in a new generation of (overly) distrustful managers. Considering the perspective of organizations, it could be that they are accidently screening for low-trust individuals, more so among younger cohorts than among older cohorts. An example is that, compared to older-generation managers, younger-generation managers are more often screened for having participated in some management training programs and that such participation is more likely among low-trust individuals than among high-trust individuals. Overall, an interesting avenue for future research is to analyze how managers are selected (or not) and explicitly link forces of selection on the one hand to the observed distinct generational trust trends and managerial generalized trust gap on the other.

This article has further studied managers' generalized trust and not particularized trust such as the trust that managers have in their supervisees. This focus was deliberate but also means that an interesting avenue for future research remains, namely to check how much of the observed generational decline is present in various other forms of intra-organizational trust (Dietz and Den Hartog, 2006). I should note, though, that data requirements are likely prohibitive. As alluded to above, separating genuine cohort effects from confounding age and period effects requires trust data collected at multiple time points, the more time points, the better. Hence, collecting the required data will be difficult and involve substantial resources.

There has been much debate about the rise of managerialism and how it is altering managers' view and treatment of other organizational members such as increasing their distrust in employees. Paralleling this debate, other fields have been studying the extent to which people's trust in most other people, i.e. generalized trust, has been declining in many different societies worldwide. Trust is widely recognized for its positive contribution to both organizational and societal outcomes, making these trends relevant for HR scholars and practitioners alike.

This article finds a significant downward generational shift in managers' generalized trust. I find a similar negative trend among the general population (employees/non-managers). However, the downward trend is much more pronounced among managers. The observed generational shifts, and the comparative strength of this shift among managers vis-à-vis non-managers, cannot be attributed to systematic differences between older and younger managerial cohorts involving, among others, these cohorts' trust in institutions, religiosity or field of study. Hence, I conclude that the idea that managers' distrust is rising is not just popular lore but supported by empirical evidence. At the same time, HR professionals are advised to prepare for a new breed of managers that is less trusting of most people than likely is optimal for their organizations.

Ethical approval was not required for this study because it relied exclusively on publicly available, de-identified secondary data.